Datasets › USPTO-50k

USPTO-50k

Introduced by Nadine Schneider et al. in What’s What: The (Nearly) Definitive Guide to Reaction Role Assignment22 Nov 2016 archive 2025-07-28

Subset and preprocessed version of Chemical reactions from US patents (1976-Sep2016) by Daniel Lowe. It includes 50K randomly selected reactions that was later classified into 10 reaction classes by Nadine Schneider et al.

Benchmarks archive 2025-07-28

All 1 leaderboard whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

First row (archive order)PaperCode
Single-step retrosynthesis USPTO-50k NAG2G (reaction class as prior) Top-1 accuracy 67.2 Node-Aligned Graph-to-Graph (NAG2G): Elevating... dptech-corp/nag2g 35 Compare

Papers archive 2025-07-28

21 shown of 21 papers with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 60. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

DateSamples run Syntology
UAlign: Pushing the Limit of Template-free Retrosynthesis Prediction with Unsupervised SMILES Alignment 1 2 25 Mar 2024 not harvested
Node-Aligned Graph-to-Graph (NAG2G): Elevating Template-Free Deep Learning Approaches in Single-Step Retrosynthesis 1 2 27 Sep 2023 ran 3 of 3 samples (0 unverified; 3 pointer-only for licence)
O-GNN: Incorporating Ring Priors into Molecular Modeling 1 2 1 May 2023 not harvested
MARS: A Motif-based Autoregressive Model for Retrosynthesis Prediction 0 2 27 Sep 2022 not harvested
G2GT: Retrosynthesis Prediction with Graph to Graph Attention Neural Network and Self-Training 1 1 19 Apr 2022 not harvested
SemiRetro: Semi-template framework boosts deep retrosynthesis prediction 0 2 12 Feb 2022 not harvested
Chemformer: a pre-trained transformer for computational chemistry 1 1 31 Jan 2022 not harvested
Improving Few- and Zero-Shot Reaction Template Prediction Using Modern Hopfield Networks 1 1 15 Jan 2022 not harvested
RetroComposer: Composing Templates for Template-Based Retrosynthesis Prediction 1 2 20 Dec 2021 not harvested
Permutation invariant graph-to-sequence model for template-free retrosynthesis and reaction prediction 1 2 19 Oct 2021 ran 0 of 6 samples (6 unverified)
RetroPrime: A Diverse, plausible and Transformer-based method for Single-Step retrosynthesis predictions 1 1 15 Sep 2021 not harvested
Deep Retrosynthetic Reaction Prediction using Local Reactivity and Global Attention 1 2 5 Aug 2021 not harvested
Learning Graph Models for Template-Free Retrosynthesis 0 2 4 Jun 2021 not harvested
Energy-based View of Retrosynthesis 0 2 14 Jul 2020 not harvested
Molecule Edit Graph Attention Network: Modeling Chemical Reactions as Sequences of Graph Edits 1 2 27 Jun 2020 ran 0 of 6 samples (6 unverified)
A Graph to Graphs Framework for Retrosynthesis Prediction 0 1 28 Mar 2020 not harvested
State-of-the-Art Augmented NLP Transformer models for direct and single-step retrosynthesis 1 1 5 Mar 2020 not harvested
Retrosynthesis Prediction with Conditional Graph Logic Network 1 1 6 Jan 2020 ran 0 of 4 samples (4 unverified)
Molecular Graph Enhanced Transformer for Retrosynthesis Prediction 0 3 25 Sep 2019 not harvested
Computer-Assisted Retrosynthesis Based on Molecular Similarity 1 2 16 Nov 2017 not harvested
Retrosynthetic reaction prediction using neural sequence-to-sequence models 1 1 6 Jun 2017 not harvested

Dataset loaders archive 2025-07-28

1 loader as listed in the archive; links are outbound and not re-checked here.

Tasks archive 2025-07-28

License archive 2025-07-28

No licence recorded in the archive. Absence here is not a statement about the dataset's terms.

Modalities archive 2025-07-28

No modality tagged.

Languages archive 2025-07-28

Variants archive 2025-07-28

  • USPTO-50k

1 variant name, as the archive lists them.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections